risk-aware path planning

This work addresses the high-risk motion planning challenges in planetary environments arising from terrain mechanics and perception uncertainties by proposing a safety-aware planning framework that integrates risk awareness with dynamical feasibility. The approach leverages AO-RRT to generate asymptotically cost-optimal initial trajectories and employs sequential convex programming (SCP) to solve the resulting nonlinear optimization problem. Notably, it introduces Conditional Value-at-Risk (CVaR) into planetary motion planning for the first time, enabling quantitative assessment and optimization of trajectory risk. Experimental results demonstrate that the proposed method reduces trajectory risk by over 97% in both simulation and hardware platforms, significantly enhancing navigation safety and mission reliability for autonomous robots operating in unknown planetary terrains.

This work addresses the challenge of ensuring recursive feasibility in safe motion planning under uncertain and time-varying environments. The authors propose a Probabilistic Recursively Feasible Model Predictive Control (PRF-MPC) framework that, for the first time, provides rigorous probabilistic guarantees of recursive feasibility. The key innovation lies in introducing the notion of “distributional consistency,” which enables the construction of an ideal predictor and yields closed-form expressions for the mean and covariance of future trajectories. Building on this, the method designs probabilistic safety constraints that, with high probability, ensure the current safe set is contained within future safe sets. Simulations in a lane-changing scenario demonstrate that the approach significantly enhances recursive feasibility, validating its effectiveness and robustness in dynamic, uncertain environments.

This paper addresses collaborative motion planning for homogeneous linear multi-agent systems operating in unknown obstacle-rich environments without explicit system models. Method: We propose a fully data-driven framework that is dynamically feasible and provably safe. It learns feedback gains and local invariant ellipsoids—serving as safety certificates—by solving a semidefinite program on experimental data. Distributed, optimization-free trajectory generation is achieved by integrating grid-based RRT sampling with a spatiotemporal resource reservation mechanism. Contribution/Results: To the best of our knowledge, this is the first work to unify data-driven invariant set learning with spatiotemporal reservation. Relying solely on limited experimental data and convex optimization tools, it simultaneously guarantees collision avoidance with static/dynamic obstacles and inter-agent collisions. The framework significantly reduces computational overhead while providing formal safety guarantees. Extensive simulations validate its effectiveness under tight dynamical constraints and complex obstacle configurations.

In dense dynamic environments, real-time path planning faces challenges including high prediction computational overhead and coarse risk assessment. This paper proposes HyPRAP, a risk-aware hybrid path planning framework. First, it introduces a Prediction-based Collision Risk Index (P-CRI) to dynamically identify high-risk obstacles and allocate high-fidelity prediction models accordingly. Second, it employs multi-model conformal prediction to generate joint confidence bounds, enabling uncertainty quantification and synergistic optimization of computational resources guided by risk. Third, it integrates Model Predictive Control (MPC) for efficient real-time trajectory generation. Simulation results demonstrate that HyPRAP significantly reduces computational load while maintaining safety; P-CRI achieves higher risk discrimination accuracy than conventional distance-based metrics; and the overall framework outperforms single-predictor baselines in the safety–efficiency trade-off.

To address the slow generation and low reliability of convex sets in configuration space for real-time robotic motion planning under dynamic environments, this paper proposes the first GPU-accelerated online probabilistic collision-free convex decomposition method—Safe Convex Sets (SCS). Our approach enables efficient iterative refinement of SCS sequences via parallelized configuration-space inflation, joint SCS optimization, trajectory-guided collision-feedback pruning, and Dynamic Random Map (DRM) search. Furthermore, we integrate piecewise-linear path inflation with nonlinear trajectory optimization subject to convex-set constraints to support perception-closed-loop online planning. Evaluated on standard simulation benchmarks, our method achieves a 17.1× speedup over CPU-based baselines and improves collision-free success rate by 27.9%. Real-world experiments on a KUKA iiwa 7 robot demonstrate millisecond-level response times and high robustness in dynamic settings.

This work proposes a hybrid navigation architecture for autonomous surface vehicles operating in dynamically uncertain marine environments, where conventional methods struggle to balance safety and real-time performance. The framework integrates a probabilistic risk map, a risk-aware RRT* planner, and B-spline-based trajectory optimization. The planner supports three reconnection modes—shortest path, minimum risk, and composite optimization—combined with a multi-objective cost function to enable dynamic adaptation to environmental hazards. Experimental results demonstrate that the proposed system significantly enhances navigational safety and trajectory smoothness in scenarios involving both static and dynamic obstacles, outperforming traditional LIDAR- or vision-only navigation approaches.

This work proposes an online trajectory generation method based on piecewise quintic/quartic splines to address the challenge of converting arbitrary geometric paths into kinematically feasible and collision-free trajectories in dynamic environments. The approach explicitly enforces jerk constraints and supports real-time replanning under high-frequency goal updates. By integrating dynamic environment perception and a responsive adaptation mechanism, it guarantees collision avoidance within finite time while permitting bounded deviations from the original path. Both simulation and real-world experiments demonstrate that the method outperforms existing approaches in trajectory smoothness, computational efficiency, and real-time performance, achieving stable operation in human-in-the-loop dynamic scenarios with target update rates up to 1 kHz.

This work addresses the challenge of efficiently generating robot arm trajectories with formal safety guarantees in non-convex environments under motion and environmental uncertainties. The authors propose a risk-bounded motion planning framework that models non-Gaussian state distributions using a Rigid-body Moment-based Deep Stochastic Koopman Operator (RM-DeSKO), integrates Sum-of-Squares (SOS) programming for binary collision certification, and introduces a hierarchical parallel verification mechanism that fuses physics-based simulation with SOS analysis to enable fine-grained collision risk quantification. This certified safety assessment is then embedded into a Model Predictive Path Integral (MPPI) controller, achieving— for the first time—provably safe trajectory optimization and sim-to-real transfer under non-Gaussian uncertainty. Experiments on two robotic manipulators and human-robot collaboration scenarios demonstrate the method’s safety, computational efficiency, and generalization capability.

This work addresses the challenge of safe and real-time trajectory planning in three-dimensional, dynamically uncertain environments, where conventional methods often fail to balance safety and computational efficiency. The authors propose SANDO, a novel system that first computes a global path using heatmap-based A*, then constructs a time-layered spatiotemporal safety corridor by inflating obstacles only with their worst-case reachable sets at each time step. Trajectory optimization is performed via mixed-integer quadratic programming (MIQP) with hard constraints, accelerated significantly through variable elimination. The approach provides formal collision-avoidance guarantees, achieving the highest success rate without constraint violations in both static and densely dynamic scenarios, while accelerating optimization by up to 7.4×. Real-world drone experiments demonstrate robustness under perception-closed-loop conditions, with 16 successful and safe flights.

This work addresses safe navigation in dynamic environments with uncertain, time-varying obstacles by anticipating local observations. It introduces the first integration of precise contingency planning with Safe Interval Path Planning (SIPP) to generate formally verified safe macro-actions. The approach performs bounded AND/OR search over a cached action–observation graph to select optimal action sequences for each reachable observation. To guide search efficiently, it employs optimistic and robust SIPP relaxations that yield admissible heuristic bounds. Decisions are made dynamically based on local observations, enabling real-time adaptation. Experiments demonstrate superior performance over fixed-path baselines in controlled road networks and successful planning in gated scenarios where conservative methods fail. The study also reveals a scalability bottleneck as observation uncertainty increases.